Objective evaluation of existing and proposed decompression procedures as well as further insights into the complete and precise etiology of cecompression sickness await optimal use of bubble counting techniques. These techniques, now accepted in diving research and medicine, still suffer from qualitative and quantitative uncertainties. Their optimal use depends on further improvements of signal processing in order to insure that all bubble signals are counted. All current Doppler studies depend partly, or completely, on the human ear. While the latter remains reliable at low (under 100 bubbles per minute) counting rates, higher rates require electronic techniques to provide acceptable count capture. This situation prevents the use of Doppler techniques in any but a semi-quantitative fashion. The object of this proposal is to increase the accuracy and reliability of gas bubble quantification by 1) reducing the amount of false counts generated by Doppler flow noise, 2) providing compensation for signal level changes, and 3) improving operator-system interaction. This will be done by further development of an existing computer algorithm processing the output of the Applied Physics Laboratory bubble counter currently in use at VMRC. The feasibility of the approach has already been demonstrated in preliminary work; what is now needed is an expanded, more comprehensive analysis from which to determine the required signal processing methods. This will allow more consistent and quantitative analysis of past and future Doppler signal recordings. Variable responses of individuals to a given decompression table require that any physical technique used in such studies remain completely reliable and reproducible from laboratory to field situation. The proposed work will provide a method with this degree of objectivity.